Igor Grigorov, RK3ZK [reference article entitle3d “The Broadband Vertical Asymmetrical Antenna” in the issue of A

VK7RO MaterialDocument2001HISTORICSPARKBelrose

Recognised text

Text below was produced by optical character recognition and may contain errors. Check it against the original.

ON MULTI-WIRE FAN MONOPOLE ANTENNAS
Igor Grigorov, RK3ZK [reference article entitled
“The Broadband Vertical Asymmetrical Antenna”
in the November 2001 issue of antenneX] has
revisited multi-wire fan antenna systems. As
background he references two articles written by
Matiychenko, UW4HW and John Schultyz, W2EEY,
published in popular radio amateur literature in 1968.
But multi-wire fan antennas have been used much
earlier, dating in fact to the turn of the last century.
For example Marconi used a 54-wire fan with his
sender at Poldhu, Cornwall, for his first transatlantic
experiment in December 1901 --- his receiver was
located on Signal Hill, Newfoundland.

I have experimentally and numerically modeled
2-, 3-, 5-, 7-, 12-, 24- and 54-wire fans --- fan shape
similar to the antenna system used by Marconi Figure 1 Igor’s Sketch of a 9-Wire Fan monopole type
[John S. Belrose, “Revisiting Wire Aerials at the antenna with wooden support poles.
Turn of the Century”, Proceedings of the Fourth
International Symposium on Antennas and EM
Theory, Xi’an, China, 1997, pp. 73-78]. The purpose
of this study was to determine the frequency used for
Marconi’s first transatlantic experiment*.

A multi-wire fan antenna, see Figure 1, with the ends
of the sloping wires connected at the top, is a much
more complicated antenna system than suggested by
what Igor has written. And, it is not a broadband
antenna, at least the wire versions that I have studied.
The antenna’s input impedance vs. frequency
response is complicated by the fact that both antenna
mode and transmission line mode currents can flow
on the structure. The transmission line mode currents
flow around the 8-triangular loops formed by the
wires of differing lengths.

If the ends of the sloping wires are not connected,
the antenna response is simple. Resonance is
determined by the lengths of the sloping wires --- and
an antenna can easily be constructed that will
resonate on several amateur bands (as Igor has
suggested). The only complication is that some
frequencies may excite resonance in two wires, e.g. a
21 MHz frequency will excite a resonance in a wire
dimensioned for this frequency, as well in a wire
dimensioned for 7 MHz, since this wire is 3/4
wavelength long for the 21 MHz frequency.

For this Technical Note (TN) I have modeled (using
NEC-4D) the 9-wire fan (see Figure 1) dimensioned Figure 2 SWR vs. frequency (5 – 27 MHz) for Figure 1’s
as in Igor’s article. To determine the radiation antenna, for two source impedances: 15 ohms and 200
characteristics I have assumed for this TN a PEC ohms. Note: PEC ground.
ground. The SWR vs. frequency response (see
Figure 2 upper curve) implies a resonance (low vertically polarized antennas over real ground the
SWR) at only one frequency (6 MHz) for Zsource = 15 efficiency is particularly poor if the take-off angle
ohms (that is Zsource low). decreases to 10o or less. This fan antenna is therefore
not a very useful radiator at frequencies near the high
The inductive reactance of the antenna’s impedance end of the band. At 27 MHz the take off angle is 6o
is low at 6 MHz and 19.8 MHz (see Table 1). Hence for seawater conductivity. For average ground in
if Zsource = 200 ohms, the SWR is low at 19.8 MHz front of the antenna (5 mS/m, 13) the low angle lobe
(the lower curve in Figure 2). is heavily attenuated, and the high angle lobes
dominate the vertical plane pattern (see Figure 3).
Table 1
Impedance vs. frequency for Figure 1’s antenna over a
PEC ground.

Frequency Impedance
MHz Ra + j X a
6 14 + j 1.4
8 28 + j 59
10 55 + j 111
12 100 + j 154
14 171 + j 172
16 214 + j 107
19.8 193 + j 6
22 123 + j 8
24 71 + j 49
26 53 + j 114

Sharply resonant responses are found at 15.7, 19.1,
21.8 and 24.4 MHz --- see Figure 2. These sharp
resonant responses are the transmission line mode
resonances referred to above.

The antenna appears to be broadly anti-resonant (high
SWR) at frequencies between 6 and 20 MHz (see
upper curve in Figure 1), but the reactive component
of the input impedance does not pass through zero.
The antenna is inductive (according to NEC-4) for all
frequencies > 6 MHz, see Table 1.

The azimuthal radiation pattern (not shown)
changes from an omni-directional pattern for
frequencies near the low end of the band, to Figure 3 Principle plane, vertical plane patterns for 6- ,
broadside directivity at 20 MHz. At 27 MHz the 19.8- and 27 MHz for a well grounded antenna with sea
directivity is a maximum in the plane of the antenna water (black maximum gain trace) and average ground (red
trace) in front of the antenna.
(minor lobes in the orthogonal directions).

The vertical plane principle plane radiation
Finally, a word of caution for antenna modellers:
pattern is cosinusoidal at the low end of the band.
MININEC and NEC-4 can have a problem in
High angle lobes start to develop at a frequency of
modelling a single wire feed feeding many wires (the
about 21 MHz. While the principle pattern lobe is
source is on the single wire). And, NEC-2 should not
directed toward the horizon for the above range of
be used, since at the junction NEC-2 considers only
frequencies (5 – 27 MHz) for very very good ground
the radius of each wire. This program does not take
conductivity (sea water) in front of the antenna, the
into account the charge distribution due to currents
elevation angle for maximum gain (take-off angle)
flowing on adjacent wires. NEC-4 and MININEC use
decreases rapidly with increase in frequency, for
a quasistatic approximation taking account of
frequencies near the high end of the band. For
neighboring wires, so should do better. My study did and downward on the wires forming the cone, excite
not establish guidelines for modelling multi-wire resonance when the vertical distance between the disc
fans. All I can say is that I found excellent agreement and the cone at a particular frequency is a quarter
between my experimental and numerical models (for wavelength. This antenna can have a 15:1 bandwidth
both antenna mode and transmission line mode (low SWR referenced to 50 ohms), but the low angle
responses) for fans comprising only a few wires. The lobe starts to lift off the horizon for frequencies
agreement between experiment and simulation greater than about 3:1 (bi-conical horn effect).
becomes worse with increase in frequency.

Conclusions: A multi-wire fan is not a broadband
antenna. But the maximum excursion in impedance is John S. (Jack) Belrose, VE2CV
reduced compared with a wire or a tower monopole. 28 November 2001
If one wants a vertically polarized HF or VHF/ UHF
broadband wire/or rod antenna the disc-cone
antenna is perhaps one of the better antenna types * [John S. Belrose, “A Radioscientist,s Reaction to
[John S. Belrose, "The HF Discone Antenna", Marconi,s first transatlantic Wireless Experiment –
QST, July 1975, pp. 11-14, 56]. The broad Revisited”, IEEE APS/URSI Meeting, Boston, MA,
bandwidth results from the radial symmetry in the July 2001].
antenna system. Currents flowing outward on the disc

~ antenneX – December 2001 Online Issue No. 56 ~